- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
intel and others are about to hit the physical limits of silicon, so what's next
Posted on 5/28/14 at 8:18 pm
Posted on 5/28/14 at 8:18 pm
intel is apparently having enough issues with manufacturing transistors at a 14nm scale, and more problems will only arise when they continue to push the limits of silicon.
so I mean...we're only like 5-7 years away from hitting the wall with silicon, so what else is there? graphene still seems like a lab project and it's extremely unlikely that it's going to be on consumer devices by 2020.
what at this point is most likely to replace silicon?
so I mean...we're only like 5-7 years away from hitting the wall with silicon, so what else is there? graphene still seems like a lab project and it's extremely unlikely that it's going to be on consumer devices by 2020.
what at this point is most likely to replace silicon?
Posted on 5/28/14 at 8:26 pm to demosa
Graphene Valley doesn't have the same ring to it.
Posted on 5/28/14 at 8:42 pm to ellunchboxo
quote:
Graphene
Beat me to it...its the fixall of the future..
Posted on 5/29/14 at 12:28 am to demosa
It has to be graphene but I have a buddy doing a PhD on some properties of graphene and he seems to believe we are pretty far off from having graphene based transistors. I know guys in a research group working on carbon nanotube transistors but I think the yield is still way too low to be cost effective.
Posted on 5/29/14 at 12:43 am to demosa
quote:
so I mean...we're only like 5-7 years away from hitting the wall with silicon, so what else is there? graphene still seems like a lab project and it's extremely unlikely that it's going to be on consumer devices by 2020.
True, but hardware has far outpaced software in terms of effective use of resources. Even if processing unit development stopped there would still be advances of performance.
Also, there is a lessening market drive at the moment for more processing power. Outside of a few exceptions, you only really utilize the power of an upper mainstream processor with multitasking.
Posted on 5/29/14 at 9:59 am to Volvagia
quote:
Also, there is a lessening market drive at the moment for more processing power. Outside of a few exceptions, you only really utilize the power of an upper mainstream processor with multitasking.
And they have other ways to improve performance.
The real game is in low power chips, and intel knows it.
Posted on 5/29/14 at 10:29 am to demosa
[link=(gizmodo.com/this-brain-inspired-microchip-is-9-000-times-faster-tha-1569176926)]LINK[/link]
This looks like we still have a ways to go with silicon.
ETA: I didn't read the title completely, was up all night.
This looks like we still have a ways to go with silicon.
ETA: I didn't read the title completely, was up all night.
This post was edited on 5/29/14 at 10:35 am
Posted on 5/30/14 at 9:26 am to eScott
quote:
ETA: I didn't read the title completely, was up all night.
You also posted a bad link.
Posted on 5/30/14 at 12:58 pm to foshizzle
Intel is not the only player. There is also TSMC and the IBM Global Foundry alliance.
But yes...we are starting to hit the wall.
But yes...we are starting to hit the wall.
Posted on 6/1/14 at 11:47 am to demosa
quote:
intel and others are about to hit the physical limits of silicon, so what's next
I think we always knew that we would reach this point some day. After all the theoretical limit for the transistor gate length would have been a single atom when it is all said and done. We have been heading towards this point (just like the end of oil) for some time now.
Having said let me ask some fundamental questions:
Does it really matter if we hit a physical wall in terms of how small transistors/chips can get? Everything that needs a processor that needs to be small (smartphones, smart watches etc.) are already small enough. The main driver once the physical limit is reached would be to make things more efficient to reduce power consumption and increase battery life. For devices that do need more power, processors can always be stacked for parallel processing - 2, 4, 8 whatever.
So in essence what I am saying is, does the end of Moore's law even have effect on normal day to day activities of a lay person? I am not so sure it does.
Posted on 6/1/14 at 1:31 pm to euphemus
quote:
Does it really matter if we hit a physical wall in terms of how small transistors/chips can get? Everything that needs a processor that needs to be small (smartphones, smart watches etc.) are already small enough. The main driver once the physical limit is reached would be to make things more efficient to reduce power consumption and increase battery life. For devices that do need more power, processors can always be stacked for parallel processing - 2, 4, 8 whatever.
If think that is the whole point though.
It's not to miniaturize further.
Small architectures allow for greater parallel processing. This has been the trend for quite some time now...air cooled processors have long since hit their speed bumps.
If you can't shrink it further, then adding more processors is not realistic. It just rapidly gets too big.
Posted on 6/2/14 at 2:07 am to Volvagia
I guess what I am saying is for hardware such as desktops/servers that require large amounts of computing power, you can still use multiple processors to get the job done, since they don't have size limitations. For wearables, smart phones, tablets etc. on which you want to watch 4K videos at 500-600 ppi or whatever, I'm sure processors at that time will be able to handle that. Beyond that, at which point the human eyes can't tell the difference, what is the point of increasing specs for specs change?
The closest analogy I can come up with is this: The world is running out of oil and gas is at a premium and everyone's focus is to maximize fuel economy barring a few specialized vehicles for certain use cases. if the max speed you can go on the freeway is limited to 65 mph at the time, what is the point of building a car with 400 bhp with a top speed to 155 mph in those circumstances? As we hit the Moore's law wall, I see the same with semiconductors - existing silicon tech will be good enough for most mobile consumer applications with a focus on efficiency and battery life, while the heavy lifting will be done by desktops and mainframes (think cloud server - thin client).
The closest analogy I can come up with is this: The world is running out of oil and gas is at a premium and everyone's focus is to maximize fuel economy barring a few specialized vehicles for certain use cases. if the max speed you can go on the freeway is limited to 65 mph at the time, what is the point of building a car with 400 bhp with a top speed to 155 mph in those circumstances? As we hit the Moore's law wall, I see the same with semiconductors - existing silicon tech will be good enough for most mobile consumer applications with a focus on efficiency and battery life, while the heavy lifting will be done by desktops and mainframes (think cloud server - thin client).
This post was edited on 6/2/14 at 2:08 am
Posted on 6/2/14 at 2:59 am to euphemus
I got that.
I didn't disagree with it. In fact, I basically said the same earlier in the thread.
I was just saying that this issue isn't about being able to make processors smaller for mobile applications as suggested by your first post.
I didn't disagree with it. In fact, I basically said the same earlier in the thread.
I was just saying that this issue isn't about being able to make processors smaller for mobile applications as suggested by your first post.
Popular
Back to top
Follow TigerDroppings for LSU Football News